Adam,Jonathan
Now using the option all for -G content and -I content , the core file
is much bigger and ::findleaks is working perfectlly.
# coreadm -h
coreadm: illegal option -- h
usage:
coreadm [ -g pattern ] [ -i pattern ] [ -G content ] [ -I content ]
[ -e {global | process | global-setid | proc-setid | log} ]
[ -d {global | process | global-setid | proc-setid | log} ]
coreadm [ -p pattern ] [ -P content ] [ pid ... ]
coreadm -u
#
# coreadm
global core file pattern: /software/rtpcore/core.%f.%p
global core file content: all
init core file pattern: /software/rtpcore/core.%f.%p
init core file content: all
global core dumps: disabled
per-process core dumps: enabled
global setid core dumps: enabled
per-process setid core dumps: enabled
global core dump logging: enabled
#
> ::findleaks -dv
findleaks: maximum buffers => 5538
findleaks: actual buffers => 5167
findleaks:
findleaks: potential pointers => 476621388
findleaks: dismissals => 467057088 (97.9%)
findleaks: misses => 9348415 ( 1.9%)
findleaks: dups => 210725 ( 0.0%)
findleaks: follows => 5160 ( 0.0%)
findleaks:
findleaks: elapsed wall time => 32 seconds
findleaks:
BYTES LEAKED VMEM_SEG CALLER
8192 7 fffffd7f937b7000 MMAP
16384 1 fffffd7ffdf8d000 MMAP
4096 1 fffffd7ffdee6000 MMAP
4096 1 fffffd7ffdab8000 MMAP
4096 1 fffffd7ffd1a9000 MMAP
4096 1 fffffd7fa572e000 MMAP
4096 1 fffffd7fa0ebe000 MMAP
------------------------------------------------------------------------
Total 7 oversized leaks, 45056 bytes
CACHE LEAKED BUFCTL CALLER
----------------------------------------------------------------------
Total 0 buffers, 0 bytes
mmap(2) leak: [fffffd7f937b7000, fffffd7f937b9000), 8192 bytes
mmap(2) leak: [fffffd7ffdf8d000, fffffd7ffdf91000), 16384 bytes
mmap(2) leak: [fffffd7ffdee6000, fffffd7ffdee7000), 4096 bytes
mmap(2) leak: [fffffd7ffdab8000, fffffd7ffdab9000), 4096 bytes
mmap(2) leak: [fffffd7ffd1a9000, fffffd7ffd1aa000), 4096 bytes
mmap(2) leak: [fffffd7fa572e000, fffffd7fa572f000), 4096 bytes
mmap(2) leak: [fffffd7fa0ebe000, fffffd7fa0ebf000), 4096 bytes
>
Thanks! So much for your help , I really apreciate it.
Every day we are finding a lot problems using MDB and DTRACE.
Valdemar
-----Original Message-----
From: ext Adam Leventhal [mailto:a...@eng.sun.com]
Sent: Wednesday, September 03, 2008 3:53 PM
To: Jonathan Adams
Cc: Pavesi, Valdemar (NSN - US/Boca Raton); mdb-discuss at opensolaris.org
Subject: Re: [mdb-discuss] How I can interpret it : couldn't read xxx
bytesno mapping for address
> If you send me the output of:
>
> % pmap partial_core > output
> % echo "::umem_debug; ::findleaks -v" | mdb partial_core >>
output
>
> I can investigate.
Hey Jonathan,
I have the core file, and will send you the location privately. Also,
it's
not a partial or incomplete core.
Adam
--
Adam Leventhal, Fishworks http://blogs.sun.com/ahl
